so-called Meta-Elements. 115 



researches is that inferences drawn from spectrum analysis 

 per se are liable to grave doubt, unless at every step the 

 spectroscopist goes hand in hand with the chemist. Spectro- 

 scopy may give valuable indications, but Chemistry must after 

 all be the final court of appeal." 



I think it has been shown that, at least in the case of the 

 yttria earths, Chemistry has not been made the final court of 

 appeal, or if it has it has signally failed to prove its case. 



The existence of meta-elements has been sought to be 

 established indirectly in ways other than that of the frac- 

 tionation of the rare earths. Thus A. E. Nordenskiold 

 (Comptes Rendus, November 2nd, 1886) finds that the earth 

 gadolinia, which consists of a mixture of yttria, erbia, and 

 ytterbia, has, though palpably a compound body, a constant 

 atomic weight whatever be the mineral from which it has 

 been extracted. " Therefore," says he, " oxide of gadolinium, 

 though it is not the oxide of a simple body but a mixture of 

 three isomorphous oxides (even when it is derived from totally 

 different minerals found in localities far apart from each 

 other), possesses a constant atomic weight. We are then in 

 the presence of a fact altogether new in Chemistry ; for the 

 first time we are confronted with the fact that three isomor- 

 phous substances, of a kind that chemists are still compelled 

 to regard as elements, occur in nature not only always 

 together but in the same proportions." 



This may be explained by saying that these three earths 

 form a compound together, just as we form compounds of 

 alumina with silica and lime or magnesia in the form of com- 

 plicated silicates, which, being chemical compounds, have a 

 constant composition whatever be the locality from which 

 they are obtained. Surely Nordenskiold does not mean to 

 state that none of the earths yttria, erbia, or ytterbia ever 

 occur except together, and that too in perfectly constant pro- 

 portions? Has not Crookes quite recently shown that 

 yttrium is one of the most widely distributed of elements ? 

 He does not mention that erbia and ytterbia also always 

 occur together with this yttria, and that too always in the 

 same proportions. Also, do not minerals as samarskite con- 

 tain many other earths besides the three in question ? 



It has thus been shown that the resolution of yttria into 

 several other new earths has by no means been proved. May 

 we not reasonably expect that as further investigation is made 

 in the subject it will be found that these different radiant- 

 matter spectra are only due to the combinations of two or 

 perhaps three earths already known and recognized as oxides of 



